{ "total_labeled_cases": 393, "total_unlabeled_cases": 588, "1a) QA model answer rate in top 3": 0.944, "1b) QA model okay rate in top 3": 0.019, "1c) QA model bad rate in top 3": 0.037, "total labeled in top 3": 107, "1a) QA model answer rate in top 5": 0.902, "1b) QA model okay rate in top 5": 0.037, "1c) QA model bad rate in top 5": 0.061, "total labeled in top 5": 163, "2) QA model ndcg metric": 0.892, "3a) QA model bad answer placement error - median": 6.944, "3b) QA model bad answer placement error - 75th percentile": 10.75, "4a) QA model answer placement error - median": 2.545, "4b) QA model answer placement error - 75th percentile": 3.5 }